78237 2012-02-09 07:25:07 -0800 [Qt][ARM] REGRESSION(r107126): It broke zillion tests 2013-06-13 03:02:21 -0700 1 1 1 Unclassified WebKit JavaScriptCore 528+ (Nightly build) All All RESOLVED FIXED Qt, QtTriaged P1 Critical --- 78145 79666 1 ossy webkit-unassigned barraclough loki oliver ossy pvarga rgabor zherczeg oldest_to_newest 553144 0 ossy 2012-02-09 07:25:07 -0800 See http://build.webkit.sed.hu/builders/ARMv5%20Linux%20Qt%20Release%20%28Test%29/builds/5708 for details 565659 1 oliver 2012-02-27 08:14:01 -0800 What endianness does the Qt ARM bot use? 566514 2 rgabor 2012-02-28 02:10:20 -0800 We are using little endian (armel). (In reply to comment #1) > What endianness does the Qt ARM bot use? 566677 3 ossy 2012-02-28 07:57:47 -0800 Now tests work with a newer GCC, but it is paper overing the problem. Zoltán commented the original bug, why this change is serious, it is a time-bomb. 566841 4 oliver 2012-02-28 10:23:30 -0800 (In reply to comment #3) > Now tests work with a newer GCC, but it is paper overing the problem. Zoltán commented the original bug, why this change is serious, it is a time-bomb. Oh right, see my reply to Zoltan -- it's a bug in the ARM assembler if you're planting different sized instructions in regions we expect to be repatched. 899958 5 ossy 2013-06-13 03:02:21 -0700 It seems it works now.